[ type thesis title here ]

MODELING OF FLOW CHARACTERISTICS IN A PUMP SUMP PHYSICAL MODEL USING COMPUTATIONAL FLUID DYNAMICS

MOHD REMY ROZAINY BIN MOHD ARIF ZAINOL

UNIVERSITI SAINS MALAYSIA

2007


by

MOHD REMY ROZAINY BIN MOHD ARIF ZAINOL

Thesis submitted in fulfilment of the requirements for the degree

of Master of Science

OCTOBER 2007

ii

ACKNOWLEDGEMENTS

In the name of Allah, The Compassionate, The Merciful

The first person I would like to thank is my principal supervisor Assoc. Prof. Dr. Hj.

Ismail Abustan. His overly enthusiasm and integral view on research and his mission

for providing knowledgeable and high quality student has made deep expression on

me. I owe him lots of gratitude for having shown me this way of research. He could not

realize how much I learned from him. Besides of being of excellent supervisor, he was

also close like a father to me and other post-graduate students.

Special thanks also go to Assoc. Prof. Dr. Zulkifly Abdullah, as my co-supervisor for his

extremely knowledgeable advices, tips and encouragement that help me a lot in

staying on the right track for this research.

I owe also my sincere gratitude to Ir. Hj. Mubarak Hussein, the director of Mechanical

Services Division of Drainage and Irrigation Department (DID) of Malaysia, who gave

me opportunity to work with them and gave me untiring help during my laboratory test

in their lab. I warmly thank to Hj. Zainuddin, Mr. Lim and all technical staff of

Mechanical Services Division for their valuable expertise advised and friendly help

within my experimental study and data collection.

My appreciation and thanks is also conveying to all the academic and technical staff of

the School of Civil Engineering, Universiti Sains Malaysia for their assistance directly

and indirectly throughout the duration of this work.

iii

I would like to express my greatest gratitude and appreciation to the Intensity Research

in Priority Area (IRPA) for its financial support during my study. Without this support I

will not able to manage my financial properly.

I feel a deep sense of gratitude to my parents Mohd Arif Zainol B. Mohd Akib and

Asmah Bt. Mat who formed part of my vision and taught me the good things that really

matter in my life. To all my siblings, I’d really like to express my thanks, especially to

my brothers Mohd Hussainy and Mohd Fizry and sister, Jurainy for just being there. I

am also grateful to my lovely wife, Noor Hana Hanif Abu Bakar for being my source of

inspiration for her endless support throughout the hard time. The good taught from her

still provides a persistent inspiration for my journey in this life.

Last but not least, I wish to express my special thanks to all my fellow colleagues and

friends for their invaluable ideas throughout this work.

Wassalam

iv

TABLE OF CONTENTS

Page ACKNOWLEDGEMENTS ii

TABLE OF CONTENTS iv

LIST OF TABLES viii

LIST OF FIGURES ix

LIST OF SYMBOLS xiii

LIST OF ABBREVIATION xv

LIST OF APPENDICES xvi

LIST OF PUBLICATIONS & SEMINARS xvii

ABSTRAK xviii

ABSTRACT xix

CHAPTER ONE : INTRODUCTION

1.1 Background of the Study 1

1.2 Problem Statements 2

1.3 Objective of the Research 3

1.4 Scope of Research 3

1.5 Advantages of the Research 4

1.6 Thesis Structure 4

CHAPTER TWO : LITERATURE REVIEW

2.1

Introduction

6

2.2 Pump Intakes Structures 6

2.3 Reynolds’s Number 17

2.4 Froude Number 17

2.5 Similitude Analysis 18

2.6 Mechanism Involved in Vortex Formation 20

2.7 Free Surface and Water Interface 21

2.8 Fundamentals of Vortex Flow in Sumps 22

2.9 Problems Encountered in Pump Intakes 23

2.9.1 Free Surface Vortex 24

2.9.2 Subsurface Vortex 27

2.9.3 Prerotation and Swirl 30

2.9.4 Cavitation 31

v

2.10 Review on Computational Fluid Dynamics (CFD) 34

2.10.1 Application CFD in Physical Model of Pump Sump 35

2.11 Summary 39

CHAPTER THREE : METHODOLOGY

3.1 Introduction 40

3.2 Description of the Model 40

3.3 Dimensionless Parameter 45

3.3.1 Reynolds Number 52

3.3.2 Froude Number 47

3.4 Data Collection 48

3.4.1 Flow Measurement 48

3.4.2 Velocity Measurement 49

3.4.3 Swirl Meter Rotation 52

3.4.4 Flow Visualization 54

3.5 Summary 55

CHAPTER FOUR : COMPUTATIONAL FLUID DYNAMICS

4.1 Computational Setups 56

4.2 Simulation Procedure 56

4.3 Pre-Processing 58

4.3.1 Geometry Set-Up and Grid Generation 58

4.4 Numerical Simulation by the Solver 63

4.4.1 Solver Set-Up and Solution Control 64

4.5 Post Processing 69

4.6 Selected Planes for CFD Analysis 69

4.7 Summary 71

CHAPTER FIVE : RESULTS AND DISCUSSIONS

5.1 Introduction 72

5.2 Velocity Distribution or Measurements 72

5.2.1 Water Level of 0.3m with Flow Rate of 15L/s (Case 1) 73





vi





5.3 Swirl Meter Measurements 85

5.4 Visual Observations 87

5.5 Visual Observation with Dye 93

5.6 Discussion from Experimental Finding 95

5.7 Simulation Results and Discussion 96

5.7.1 Grid Sensitivity 96

5.7.2 Velocity Profile of Numerical Simulation 99

5.7.2.1 Water Level of 0.3m with Flow Rate of 4L/s (Case 3) 99

5.7.2.2 Water Level of 0.24m with Flow Rate of 10L/s (Case 5)

100

5.7.2.3 Water Level of 0.18m with Flow Rate of 15L/s (Case 7)

102

5.7.3 Velocity Vector (Flow Visualization) 103

5.3.3.1 Velocity Vector (Flow Visualization) at Water Level of 0.18m with Flow Rate of 15L/s (Case 7)

103


111


116

5.7.4 The Static Pressure Contour Result 120

5.8 Comparison of Visualization Results between Experimental and Simulation

122

5.9 Comparison of Velocity Profile between Experimental and Simulation Results

125

5.10 Simulation at Higher Flow Rate of 30L/s 129




5.10.4 Velocity Vector (Flow Visualization) at Water Level of 0.3m with Flow Rate of 30L/s (Case 10)

133

5.10.5 Velocity Vector (Flow Visualization) at Water Level of 0.24m with Flow Rate of 30L/s (Case 11)

139

5.10.6 Velocity Vector (Flow Visualization) at Water Level of 0.18m with Flow Rate of 30L/s

144

CHAPTER SIX : CONCLUSIONS

6.1 Conclusion 149

6.1.1 Experimental Work 149

vii

6.1.2 CFD Modelling 149

6.2 Recommendations for Future Works 151

REFERENCES 153

APPENDICES

Appendix A Components of Physical Model

Appendix B Velocity Distribution of Observed Result

Appendix C Velocity Distribution of Calibrated Result

Appendix D Velocity Distribution of Calibrated Result (30L/s)

Appendix E Display Input of FLUENTTM

Appendix F Dye Injection at Several Locations

Appendix G Velocity Vector Magnitude at Several Locations

viii

LIST OF TABLES

Page

2.1 Recommended dimensions for Figures 2.6 and 2.7 (Source: American National Standard for Pump Intake Design, 1998)

16

2.2 Free surface vortex type and classifications (Source: Larsen and Padmanabhan, 2001)

26

2.3 Types of cavitation (Source: Li, 2000) 33

3.1 The Reynolds number at different of flows at different of water levels

47

3.2 The Froude number at different of flows at different of water levels

48

4.1 The turbulence kinetic energy, k and turbulence dissipation rate, ε

59

4.2 Properties of water fluid 66

4.3 Under relaxation factors used in the simulation 66

4.4 Discritization used in the simulation 67

4.5 The reference value used in the simulation 67

5.1 Result of swirl and swirl angle of the study 85

5.2 Visual observation of vortex type and location in the pump sump

88

5.3 Comparison of visualized results between experimental and simulation

124

ix

LIST OF FIGURES

Page

2.1 Types of intake structures (Source: Knauss, 1987) 8

2.2 Guidelines for multiple pump sumps (Source: Hydraulic Institute Standards,1983)

11

2.3 Basic sump design for multiple pump sumps according to British Hydrodynamics Research Association, (a) open sump and (b) unitized sumps (Source: Prosser, 1977)

12

2.4 Basic design for a single bay sump with uniform approach flow (Source: Padmanabhan, 1987)

13

2.5 Basic design for multiple bay sumps with uniform approach flow (Source: Padmanabhan, 1987)

14

2.6 Recommended intake structure layout (Source: American National Standard for Pump Intake Design, 1998)

15

2.7 Filler wall details for proper bay width (Source: American National Standard for Pump Intake Design, 1998)

15

2.8 Vortex types - Upper sketch is Type 1, first inception; center sketch is Type 2, marginal performance; lower sketch, Type 3, starting to break suction (Source: Dicmas, 1987)

25

2.9 Classification of free-surface vortices (Source: Nakato, 2003) 27

2.10 (a) Floor vortex, (b) back wall vortex, (c) sidewall vortex and visualization of subsurface vortices (Source: Rajendran et al. 1999)

28

2.11 Classification of boundary-attach subsurface vortex (Source: Nakato, 2003)

29

2.12 General cavitation damage on the underside of a mixed flow impeller (Source: Dicmas, 1987)

32

2.13 Propeller pump damage at liner section (a) from asymmetrical flow pattern; (b) general cavitation (Source: Dicmas, 1987)

32

2.14 The multidisciplinary nature of CFD (Source: Lohner, 2001) 34

2.15 Geometrical configuration and calculated flow pattern for the intake study (Source: Guang et al. 1996)

36

2.16 Model intake and computational grid (Source: Constantinescu and Patel, 2000)

37

3.1 Location of testing laboratory of the Department of Irrigation and Drainage (DID) mechanical and electrical services (Source: Abustan et al. 2004)

40

3.2 The laboratory physical model of pump sump 41

3.3 Plan and side views of the model 42

3.4 (a) Dimensions plan, (b) profile of typical pump cell 43

3.5 Detail structure of suction intake (bell mouth) 44

3.6 Positions of the values Ainlet and Vinlet are obtained 46

3.7 Flow measurement equipment 49

3.8 Velocity measurement equipment 50

3.9 (a), (b), (c) Side view, (d) plan view of velocity measurements points

51

3.10 Four blade neutral pitch propeller 53

x

3.11 Swirl meter (vortimeter / rotometer) installed in the model column intake

53

3.12 Tachometer measurement equipment 53

3.13 Dye tracer technique for observation approach (Source: Ahmad et al. 2005)

54

4.1 Problem solving steps in the CFD analysis 57

4.2 The computer model of pump sump 59

4.3 The surface and the volume mesh of the climate chamber 60

4.4 Mesh distributed across x-y plane 61

4.5 The boundary conditions setup 62

4.6 Overview of the segregated solution method 68

4.7 Grid locations in the sump intake (upper- plan view / horizontal plane, lower- vertical plane)

70

5.1 Contour of velocity distribution of observed result in the pump sump at water depth of 0.3m and flow rate of 15L/s (Case 1)

74


75

5.3 Velocity profiles from node A to E at Row I 75


77


78


79


81


82


83


84

5.11 Visual observations of vortices occur in Case 5 89

5.12 Surface vortex Type 2 formations in the sump intake at the back of the column intake (Case 4 and 5)

89

5.13 Visual observation of vortices occur in Case 4 90

5.14 Visual observation of vortices occur in Case 8 and Case 7 91

5.15 Formation of subsurface vortex Type 2 directly under the suction intake (Case 7)

92

5.16 Shows the surface vortex Type 5 formations in the sump intake at the back of the column intake (Case 7 and 8)

93

5.17 Dye injection at the back of sump column (Case 1) 94

5.18 Dye injection in front of the sump column (Case 1) 94

5.19 Velocity profile at water level of 0.3m with flow rate of 4L/s 98



xi

5.22 Contour of velocity distribution of simulated result in the pump sump at water depth of 0.3m with flow rate of 4L/s (Case 3)

100

5.23 Contour of velocity distribution of simulated result in the pump sump at water depth of 0.24m and flow rate of 10L/s (Case 5)

101

5.24 Contour of velocity distribution of simulated result in the pump sump at water depth of 0.18m and flow rate of 15L/s (Case 7)

103

5.25 Surface vortex at the upper level of the water (0.18m) by flow rate of 15L/s

105

5.26 Surface vortex at the elevation of 0.09m at water level of 0.18m by flow rate of 15L/s (b) shows vertical plane at the elevation of 0.09m

106

5.27 Subsurface vortex at the bottom level (0m) of the sump intake at water level of 0.18m by flow rate of 15L/s

107

5.28 Subsurface vortex at the back wall of sump intake at water level of 0.18m by flow rate of 15L/s (back view)

109

5.29 Velocity vector magnitude near the wall of the sump intake at water level of 0.18m with flow rate of 15L/s (side view)

109

5.30 Contour of velocity vector at water level of 0.18m with flow rate of 15L/s (back view)

110

5.31 Contour of velocity vector at water level of 0.18m with flow rate of 15L/s (side view)

110

5.32 Two counter rotating surface vortices at the surface water level of 0.24m with flow rate of 10L/s

112

5.33 Velocity vector magnitude at the bottom of the sump intake at water level of 0.24m with flow rate of 10L/s

113

5.34 Velocity vector magnitude at the back wall of sump intake at water level of 0.24m with flow rate of 10L/s

114

5.35 Velocity vector magnitude near the wall of the sump intake at water level of 0.24m with flow rate of 10L/s

115

5.36 Contour of velocity vector at water level of 0.24m by flow rate of 10L/s (back view)

115

5.37 Velocity vector magnitude at surface area at water level of 0.3m with flow rate of 4L/s

117

5.38 Velocity vector magnitude near the wall of the sump intake at water level of 0.3m with flow rate of 4L/s

117

5.39 Flow is symmetrically directed towards the suction intake (no vortex) at water level of 0.3m with flow rate of 4L/s

118

5.40 Contour of velocity vector at water level of 0.3m with flow rate of 4L/s (back view)

119

5.41 Contour of static pressure at water level of 0.18m with flow rate of 15L/s (back view)

121


121


122

5.44 Water level of 0.3m with flow rate of 4L/s (Row 1) 126



5.47 Experimental and simulated data scatter plots at water level of 0.3m with flow rate of 4L/s (Row 1)

127


127

xii


128

5.50 Contour of velocity distribution at water depth of 0.3m and flow rate of 30L/s (Case 10)

130


131


132

5.53 Surface vortices at the surface water level of 0.3m with flow rate of 30L/s

135


136


137

5.56 Velocity vector magnitude near the wall of the sump intake at water level of 0.3m with flow rate of 30L/s (Sidewall 1)

137

5.57 Contour of velocity vector at water level of 0.3m with flow rate of 30L/s (back wall)

138

5.58 Contour of velocity vector at water level of 0.3m with flow rate of 30L/s (Sidewall 1)

138

5.59 Surface vortices at the surface water level of 0.24m with flow rate of 30L/s

140


141

5.61 Velocity vector magnitude at the back wall of sump intake at water level of 0.24m with flow rate of 30L/s (Back wall)

141


142

5.63 Contour of velocity vector at water level of 0.24m by flow rate of 30L/s (Back wall)

142


143

5.65 Two counter rotating surface vortices at the surface water level of 0.18m with flow rate of 30L/s

145


146


147


147

5.69 Contour of velocity vector at water level of 0.18m by flow rate of 30L/s (Back wall)

148


148

LIST OF SYMBOLS

xiii

ρ Density

µ Dynamic viscosity

υ Kinematics viscosity

Θ Swirl angle

ε Turbulence dissipation rate

a Inlet width

A Cross-sectional area of the intake

Ainlet Inlet cross-section

b Inlet depth

D1 Suction intake diameter (bell mouth)

D2 Column intake diameter

dv Diameter of vortimeter vane

FD Froude number

Fm Froude model parameter

Fp Froude prototype parameter

Fr Froude ratio parameter

g Gravitational constant

H Water level / depth

k Turbulence kinetic energy

l Characteristic length

L Characteristic length depending on water level

Q Flow through the pump intake

Qinlet Inlet flow rate

R Number of revolutions of the vane per minute

Re Reynolds number

S Submergence depth

TI Turbulence intensity

xiv

U Inlet velocity

V Velocity

VA Axial velocity

Vc Cross-flow velocity

Vch Approach velocity

Vinlet Inlet velocity

VR Rotational velocity

Vx Pump bay velocity

LIST OF ABBREVIATION

xv

ADV Acoustic Doppler Velocimeter

CFD Computational Fluid Dynamics

DID Department of Irrigation and Drainage

PVC Polyvinyl chloride

RANS Reynolds-average Navier Stokes

TPFM Thermo Polysonic Flow Meter

VVM Valeport Velocity Meter

3D Three dimensions

LIST OF APPENDICES

xvi

Appendix A Components of Physical Model

Appendix B Velocity Distribution of Observed Result

Appendix C Velocity Distribution of Calibrated Result

Appendix D Velocity Distribution of Calibrated Result (30L/s)

Appendix E FLUENTTM Display Input

Appendix F Dye Injection at Several Locations

Appendix G Velocity Vector Magnitude at Several Locations

LIST OF PUBLICATIONS & SEMINARS

xvii

Mohd Remy Rozainy, M.A.Z., Abustan, I., Abdullah, M.Z., Muhad, R.M. Nawi., and M.A. Ismail. (2006). Computational Fluid Dynamic (CFD) to Modeling of Flow Characteristics in Physical Model of Pump Sump. 1st Civil Engineering Colloquium, School of Civil Engineering, Universiti Sains Malaysia.

PERMODELAN CIRI-CIRI ALIRAN DALAM MODEL FIZIKAL TAKUNGAN PAM MENGGUNAKAN PERKOMPUTERAN DINAMIK BENDALIR

xviii

ABSTRAK

Kajian ini dijalankan bagi memodelkan ciri-ciri aliran dalam model fizikal takungan pam

menggunakan Dinamik Bendalir Berkomputer (CFD) melalui kod FLUENTTM 6.2.

Prosedur atau tatacara ujikaji melibatkan cerapan data mengunakan meter halaju ,

meter aliran, dan meter pusaran (Rotometer / Vortimeter). Tiga jenis pengukuran

diambil iaitu halaju, aliran, dan sudut pusaran untuk sembilan kajian kes yang telah

dinilai pada tiga kedalaman air yang berbeza (0.3m, 0.24m, dan 0.18m), masing-

masing pada tiga kadar alir yang berbeza (15L/s, 10L/s, dan 4L/s); Sejumlah 162 titik

cerapan bagi setiap kes. Suatu ujian visual yang melibatkan teknik pengesanan

penunjuk bewarna (dye) turut dijalankan untuk mencirikan aliran. Dalam kajian ini,

pembangunan model CFD yang komprehensif telah digunakan dalam rekabentuk

takungan pam. Perbandingan dengan kaedah ujikaji dan model CFD akan

dibincangkan dengan lebih lanjut. Hasilan FLUENTTM menggambarkan ciri-ciri asas

magnitud vektor-vektor halaju (m/s), kontur vektor halaju (m/s) dan kontur tekanan

pegun (pascal). Persetujuan yang baik diperolehi antara keputusan simulasi dan ujikaji.

Lokasi vortex dalam keputusan ujikaji hampir menyamai keputusan simulasi CFD yang

diperolehi dalam kajian ini. Purata perbezaan halaju di antara ujikaji dan simulasi ialah

4.2% dan 11.6%. Sementara itu, nilai pekali regresi (R2) yang berada dalam julat 0.98

ke 0.99 telah diperolehi untuk kehubungan plotan berselerak antara data ujikaji dan

simulasi. Oleh itu, daripada kajian ini, kesimpulan yang boleh dibuat ialah CFD dapat

digunakan untuk simulasi atau di masa akan datang menggantikan model fizikal

takungan pam.


xix

ABSTRACT

This study attempts to model the flow characteristic in a pump sump physical model by

using Computational Fluid Dynamics (CFD) code FLUENTTM 6.2. The experimental

procedures include the data collection using a velocity meter, flow meter and swirl

meter (Rotometer / Vortimeter). Three types of measurements were conducted which

are velocity, flow, and swirl angle for nine cases which had been evaluated at three

different water depths (0.3m, 0.24m and 0.18m) and at three different flow rates (15L/s,

10L/s and 4L/s); a total of 162 measurement points per case. A visual test that involves

the dye tracing technique was also carried out to characterize the flow. Further, in this

study, a comprehensive CFD model of pump bays was developed. The comparison of

experimental method and CFD model is discussed in details. The FLUENTTM outputs

illustrate the basic features of magnitudes of velocity vectors (m/s), contour of velocity

vector (m/s) and static pressure contour (pascal). A good agreement is determined

between simulation and experimental results. The locations of the vortices in the

experimental results closely match the CFD simulation results obtained. The average

velocity magnitude difference between experimental and CFD simulated result is

recorded at 4.2% to 11.6%. Moreover, the regression coefficient (R2) values of velocity

magnitude ranging from 0.98 to 0.99 were obtained from the scattered plot relationship

between experimental and simulated data. Thus from the study, it can be concluded

that the CFD can be used to simulate flow characteristics in pump sump as an

alternative to physical modeling of pump sump.

1

CHAPTER 1 INTRODUCTION

1.1 Background of the Study

The flow phenomena that arise in the pump bays of water intake structures

have been studied experimentally by many researchers (Dicmas, 1987; Larsen and

Padmanabhan, 2001; Ansar et al., 2002; Nakato, 2003; Tokyay and Constantinescu,

2005a) due to the importance in determining pump bays performance. Unfortunately,

the phenomena are so complex and diverse that, there is no comprehensive theoretical

model to predict them. Existing design guides, usually contains little more than rules of

thumb for pump performance. Therefore, Computational Fluid Dynamics (CFD) could

play a potentially useful role in determined flow conditions within pump bays. Improved

knowledge of flow conditions should lead to improvements in the bay design and

consequently the pumps operation. The factors affecting pump-bay flows have been

known in qualitative and empirical terms but there is no exact method for predicting

them. The only way of doing so is by an expensive hydraulic model study. Because of

the high costs involved in the design and construction of small physical scale laboratory

models, there is a need for more research in the numerical simulations. The prime

drawbacks of physical hydraulic models are the relatively lengthy periods needed for

model building, data acquisition and analysis. The numerical simulation code

introduced herein does not have those drawbacks, plus it has additional advantages. In

this study, the application of the comprehensive CFD model will be used in the design

of pump bays, and the comparison with the development experimental model will be

discussed in details.

2

1.2 Problem Statements

Water pumps in drainage, agriculture, and industrial process applications are

known to experience certain common operational problems, such as vibration, impeller

damage due to cavitation and excessive bearing wear resulting in severe deterioration

of their performance and finally lead to a significant increase of operational and

maintenance costs. These problems probably result from poor intake design.

Therefore, there are several problems that should be highlighted in this study. The

common of the main problems are summarized as follows (Larsen and Padmanabhan,

2001; Karassik et al., 2001 and Nakato, 2003):

a. Free surface vortices – the air may draw from the surface into the pump. These

types of vortices can cause unbalance loading of impeller, periodic vibration

and therefore reduction in pump capacity.

b. Subsurface vortices – which may emanate from floor, site, back walls or

combination among them. These can cause vibration and cavitation that may

reduce pump efficiency.

c. Pre-rotation – flow entering the pump which change the angle of the attack of

the impeller blades from the design value and may effect pump efficiency and

lead to cavitation.

d. An uneven distribution of flow at the pump throat which may results in unequal

loading of pump impeller. This action will lead to vibration and unbalance

loading of impeller.

e. Cavitation that can cause damage on the underside of mixed flow impeller.

These problems encountered in the pump sump will affect the pump

performance and significantly increase the operational and maintenance costs. In order

to identify sources of particular problems and find practical solution for it, the usual

approach is to conduct the laboratory experiments on a scaled physical model. The

3

problems are already there; however the solutions are still the matter. The main

concern and challenge in this study is to understand the flow characteristics. These

results will be compared with the computational result by using the Computational Fluid

Dynamic (CFD), FLUENTTM software.

1.3 Objectives of the Study

In order to accommodate the main concern and challenge stated in the problem

statement, the following objectives are set up to be find the optimum solution. These

designated objectives will serve as the basis of the problem solving and also as a

guideline and reference in order to complete the study. The objectives are as listed

below:

1. To identify flow characteristics in particular surface vortices and subsurface

vortices, velocity distribution and pressure contour at different discharge water

levels.

2. To develop a simulation model of sump intake using CFD.

3. To ascertain the effectiveness of the CFD analysis.

4. To predict the occurrences of vortices using CFD.

1.4 Scope of the Research

To achieve the above objectives designed, the following tasks need to be

carried out:

(i) Literature review – This is the first step to understand the concept and theory of the

related research area. It serves as a basic knowledge of other researchers experiences

as stated in their literature works reviewing the theory, concept and methods of their

4

studies. It is hoped that the review will ensure a good overview on the whole research

activities.

(ii) Design of the structure of physical model – The structure of the physical model

had been designed by referring to the existing design guidelines (DID, 2000). The

design of the pump sump includes the determination of material used, sump scale and

relevant components such as the approaching slope and bell mouth.

(iii) Construction of the physical model – In this part, experiences and expertise

from Drainage and Irrigation Department (DID) was utilized, by involving in construction

of a physical model. Construction of the physical model includes column intake, inlet

channel, water intake, piping system, fitting and checking valve and pumps.

(iv) Hydraulic data collection on the model testing – In this study, three types of

measurement were conducted which involve velocity, flow and swirl angle

measurements. These three measurements were conducted using different special

equipments that will be explained in the following chapter. Visual tests that engage the

dye tracing technique were also carried out to understand and identify the flow

characteristics and vortex position.

(v) Numerical simulation – The simulation had been made by using CFD code i.e.

FLUENTTM 6.2 software. The FLUENTTM model serves as a tool that discrete and

solves governing equations for specific geometries using a set of finite volume method.

(vi) Data processing, analysis, interpretation and evaluation – At this stage, all the

data from experimental methods and numerical methods have to be analyzed. Both

experimental and simulated data will be interpreted and evaluated accordingly.

5

(vii) Result assessment – This section presented results and analysis of data from

experimental physical model test, simulated by using CFD method and its comparison.

(viii) Conclusions and recommendation – It is the final chapter for this thesis, which

highlights the findings of this research and recommendations for further studies on the

related topic.

1.5 Advantages of the Research

The outcomes of the research will provide some advantages in understanding

the flow feature by experimental and numerical methods. The entire advantages are as

listed below:

1. The vortices including surface vortices and subsurface vortices can be locally

identified in the numerical model. Perhaps, other hydraulic problem such as

back flow and dead flow region can also be identified.

2. By using numerical method, in stead of construction the physical model of pump

sump, the numerical results could be utilized. These could reduce cost and

working time.

3. Operational and maintenance cost of a pump station can be cut off by curbing

the entire hydraulic problem through more efficient and effective design.

4. The developed databases can be used to remedy existing problematic pump

sump which can help other researchers.

1.6 Thesis Structure

The thesis has been categorized into specific chapters for better understanding

of the research. The lists of chapters are as follow:

6

Chapter 1: Introduction – This chapter gives an overview of the thesis including five

important things such as background of the study, problem definition, objective of the

research, scope of research and advantages of the research.

Chapter 2: Literature review and hypothesis – This chapter provides important

theoretical and conceptual understanding of related topics based on various

researches including hypothesis of the research.

Chapter 3: Experimental setup – The experimental setup of the sump intake model

will be described and it will help to fulfill the proposed designated objectives and

answer the problems defined. This chapter is the most important part in this thesis. The

experimental procedures include the data collection procedure using instrumentations.

Chapter 4: Computational Fluid Dynamics – This chapter will discuss the concept,

theory and the methods of CFD. Besides that, this chapter will give a clear view and

step by step basic understanding of CFD.

Chapter 5: Result and discussion – Results, analysis, discussion and result

assessment of experimental and simulation are described in this chapter. Comparison

of experimental and numerical methods results are described in detail. It will be

followed by analysis, discussion and result assessments. Furthermore, the result from

the CFD methods will be visualized and discussed in this chapter.

Chapter 6: Concluding remarks – The final chapter will summarize all the activities

related to this study, and all the recommendations for further works are presented here.

6

CHAPTER 2 LITERATURE REVIEW

2.1 Introduction

This chapter describes the pump intake structure, Reynolds number, Froude

number, similitude analysis, mechanism involved in vortex formation, free surface and

water interface, fundamental of vortex flow in sumps and problem encountered in the

pump intakes. Finally, a review is presented on Computational Fluid Dynamics (CFD)

and their application in the physical model of a pump sump.

2.2 Pump Intake Structures

The term ‘water intake’ refers to the channel leading from the water source

which may be a river or reservoir and all installations downstream including the pump

column or the intake pipe (the suction tube portion of a vertical intake), the approach

channel (upstream of pump bay) and the pump bay (bounded by the floor, the back

wall, side walls dividers walls separating adjacent pump column). Usually the upstream

end of the pump column has an inlet attachment called the suction bell. The function of

the intake is to supply an evenly distributed flow of water to the pump suction bell.

Intake structures can be categorized as being clear liquids or solids-bearing

liquids. For clear liquids, intakes are further classified into rectangular, formed, circular

and trench types, as well as suction tanks and cans. For solid-bearing liquids, trench

type and rectangular wet wells are usually considered. These structures are covered in

detailed laboratory studies where hydraulic modeling is frequently performed to locate

and suppress or avoid flow problems in existing water intakes. For example, Larsen

and Padmanabhan (2001) recommended model study to be undertaken for the

following water intakes conditions:

• Intakes with asymmetric approach flow (e.g., an offset in the approach channel)

7

• Intakes with multiple-pump sumps with a common approach channel and

variety of pump-operating combinations.

• Intakes with pumps capacities greater than 2.5m3/s per pump.

• Intakes with an expanding approach channel and

• Intakes with possibilities of screen blockages and/or obstructions close to the

suction-pipe entrance.

A hydraulic intake structure, such as the multiple pump sump, consists of an

open channel (pump sump or diversion channel) and a pipe or conduit. The flow in the

intake involves the transition from a free-surface flow in an open channel to a close

conduit flow in a pipe. Several of types of intake structure exist. Figure 2.1 shows the

different types of intake structures. The vertically downwards intake consists of a pipe

or conduit located just above (or near) the floor of the pump sump. Other intake

structures include a horizontal intake, inclined downward and upward intakes, and

vertically upward intakes in free-surface flow. If the flow is not driven by gravity, as in

the vertically and inclined downward intakes, a pipe is needed to withdraw water from

the pump sump to its final destination. Therefore, a pump is required in the horizontal,

the vertically upward and vertically inclined intakes. The intake that requires pumps are

commonly referred to as pump intakes. For this study, vertically upward intake is used

in the physical model of pump sump.

There are many researchers who have stated about the specific hydraulic

phenomena that can adversely affect the performance of pumps (Tullis, 1979; Dicmas,

1987; Bauer and Nakato, 1997; Larsen and Padmanabhan, 2001; ANSI, 1998 and

Warring, 1984). The hydraulic phenomenon that has been discussed are free surface

and subsurface vortices, excessive of flow entraining the pump and its variations with

time, entraining air or gas bubbles and non-uniform of velocity at the impeller eye and

excessive variations in velocity with time.

8

Figure 2.1: Types of intake structures (Source: Knauss, 1987)

The sump should be designed to allow the pumps to achieve optimum hydraulic

performance for all operating conditions. The acceptance criteria for the model test are

9

based on the Hydraulic Institute Standards (1983) recommendations and shall be as

follows:

• Free surface and subsurface vortices entering the pump must be less severe

than vortices with coherent (dye) cores (free surface vortices of Type 3 and

subsurface vortices of Type 2)

• Dye core vortices may be acceptable only if they occur for less than 10% of the

time or only infrequent pump operating conditions.

• Swirl angles, both the short-term (10 to 30 second model) maximum and the

long-term (10 minute model) average indicated by the swirl meter rotation, must

be less than 5°. Odgaard and Dlubac (1984) reported swirl rotation must be less

than 3°.

• Maximum short-term (10 to 30 second models) swirl angles up to 7 degrees

may be acceptable, only if they occur less than 10% of the time for infrequent

pump operating conditions. The swirl meter rotation should be reasonably

steady, with no rapid changes in direction when rotating near the maximum

allowable rate (angle).

• Time-average velocities at points in the throat of the bell or at the pump suction

in a piping system shall be within 10% of the cross-sectional area average

velocity. Time-varying fluctuations at a point shall produce a standard deviation

from the time-averaged signal of less than 10%.

A set of design criteria has been developed by the Iowa Institute of Hydraulic

Research (IIHR) based on their vast experience with model studies of pump sumps.

Nakato and Yoon (1992) summarize them as follows:

• No detectable boundary-attached vortices extending into the pump bells

10

• No free-surface vortices stronger than Type 2 (Arden Research Laboratory

classification)

• No velocities measured at the pump throat that vary by more than 10% from the

average of all local velocities measured in the cross section

• Vortimeter-tip velocity angles (swirl angles) no greater than 5°

• No detectable, large scale, persistent unsteadiness or waviness in the pump

bell approach flows, no indication of persistent large scale turbulence, and no

flow anomalies judged objectionable by investigators experienced with pump-

intake model test.

The guidelines listed could help to determine whether conditions for the existing

intake structures are acceptable or not. If the conditions are not acceptable,

modification to the intake structure should be made until the requirement is satisfied.

There are many guidelines or the basic designs that have been developed to improve

the reliability and performance of pump sumps. The Hydraulic Institute Standards

(1983) and British Hydrodynamics Research Association (Prosser, 1977) detailed

some recommendations for the multiple pumps in open sumps as illustrated in Figure

2.2 and Figure 2.3.

11

Figure 2.2: Guidelines for multiple pump sumps (Source: Hydraulic Institute Standards, 1983)

12

Figure 2.3: Basic sump design for multiple pump sumps according to British Hydrodynamics Research Association, (a) open sump and (b) unitized sumps (Source:

Prosser, 1977)

13

Padmanabhan (1987) has also developed guidelines for single sump and

multiple sump intakes, which are illustrated in Figure 2.4 and Figure 2.5 These

guidelines and basic sump design as discussed are considered in this study in

determining methods to minimize the vortices in the multiple pump sumps. As

Padmanabhan (1987) states, the guidelines given previously are helpful for the

preliminary design of pump sumps, but a model study should be performed for more

complex intake structures and for the evaluation of preliminary designs.

Figure 2.4: Basic design for a single bay sump with uniform approach flow (Source: Padmanabhan, 1987)

Screens

14

Figure 2.5: Basic design for multiple bay sumps with uniform approach flow (Source: Padmanabhan, 1987)

The American National Standard for Pump Intake Design (1998) recommends

that the intake design for vertical wet pit pumps is as shown in Figure 2.6 and Figure

2.7. The geometry is generally defined in terms of the pump inlet bell diameter as

shown. Once the number and the size of pump required are determined, a pump inlet

diameter can be estimated. The bell diameter can be estimated based on an inlet pipe

velocity of between 0.9m/s and 2.4m/s. As the selected bell diameter has been

determined, the proportions of the inlet structure can be estimated from Figure 2.6 and

Figure 2.7. Table 2.1 gives recommended values for the dimensions.

Select S using S/D=a+bFD with a=1-1.5 and b=2-2.5

S should also satisfy the required NPSH for the sump

15

Figure 2.6: Recommended intake structure layout (Source: American National Standard for Pump Intake Design, 1998)

Figure 2.7: Filler wall details for proper bay width (Source: American National Standard for Pump Intake Design, 1998)

16

Table 2.1: Recommended dimensions for Figures 2.6 and 2.7 (Source: American National Standard for Pump Intake Design, 1998)

Dimension Variable Description Recommended Value

A

a

B

C

D

H

h

S

W

w

X

Y

Z1

Z2 α β

Φ

Distance from the pump inlet bell centerline to the intake structure entrance Length of constricted bay section near the pump inlet Distance from the back wall to the pump inlet bell centerline Distance between the inlet bell and floor Inlet bell design outside diameter Minimum liquid depth Minimum height of constricted bay section near the pump inlet bell Minimum pump inlet bell submergence Pump inlet bay entrance width Constricted bay width near the pump inlet bell Pump inlet bay length Distance from pump inlet bell centerline to the through-flow traveling screen Distance from pump inlet bell centerline to diverging walls Distance from pump inlet bell centerline to sloping area Angle of floor scope Angle of wall convergence Angle of convergence from constricted area to bay walls

A = 5D minimum, assuming no significant cross-flow* at the entrance of the intake structure a = 2.5D minimum B = 0.75D C = 0.3D to 0.5D (see text) H = S + C

h = (greater of H or 2.5D) S = D (1.0 + 2.3FD) W = 2D minimum w = 2D X = 5D minimum, assuming no significant cross-flow at the entrance to the intake structure Y = 4D minimum. Dual-flow screens required a model study Z1 = 6D minimum, assuming no significant cross-flow* at the entrance to the intake structure Z2 = 5D minimum α = -10 to + 10 degrees β = 0 to +10 degrees(Negative values of β, if used, require flow distribution devices developed through a physical model study) Φ = 10 degrees maximum

*Cross-flow is considered significant when Vc > 0.5Vx average

17

2.3 Reynolds’s Number

In 1883, Osborne Reynolds is the first to develop the basic laws of turbulent

flow. He studied the flow of liquid in pipes and found that at a low speed the flow is

smooth but at high speed, the flow is turbulent. He found that the onset of turbulence in

a smooth pipe was related to the Reynolds number in a very interesting way. In the

case of pipe flow, where the diameter of the pipe is the characteristic length scale, for

Re < 2000 the flow is laminar, for 2000 < Re < 4000 there is a gradual change to

turbulent flow, and for Re > 4000 the flow is turbulent. In the case of open-channel

flow, such as in canals and rivers, the depth of flow is used as the characteristic length

scale and open-channel flows are turbulent for > 1000 (Chin, 2000). In most

engineering applications involving closed-conduit and open-channel flow, the Reynolds

number limits are far exceeded and the flows are fully turbulent. The formula of

Reynolds number can be calculated by the equation:

Re = = = VL VL inertial forces

viscous forcesρ

υ μ (2.1)

where,

V = velocity (m/s)

L = characteristic length (m)

υ = kinematic viscosity (m2/s)

µ = dynamic viscosity (kg/ms)

ρ = water density (kg/m3)

2.4 Froude Number

The Froude number is the ratio of inertial to gravitational forces and is defined

for pump sump as:

(2.2)

D

18

where,

FD = Froude Number

g = gravitational constant (9.81m/s)

Characteristic length in an open channel is taken to be the hydraulic depth.

Depending on the magnitude of the ratio of inertial to gravity forces, a flow is classified

as subcritical, critical or supercritical (French, 1986). If FD = 1, the flow is in a critical

state with the inertial and gravitational forces in equilibrium. If FD < 1, the flow is in a

subcritical state, and the gravitational forces are dominant. If FD > 1, the flow is in a

supercritical state and inertial forces are dominant.

When the flow is subcritical, FD < 1, the velocity of flow is less than the speed of

an elementary gravity wave. Therefore, such a wave can transmit upstream against the

flow, and upstream areas are in hydraulic communication with the downstream areas.

Furthermore, when the flow is supercritical, FD > 1, the velocity of the flow is greater

than the speed of the elementary gravity wave. Therefore, such wave cannot transmit

upstream against the flow, and the upstream areas of the channel are not in hydraulic

communication with the downstream areas. Thus, the possibility of an elementary wave

transmit upstream against the flow can be used as a criterion for differentiating

between subcritical and supercritical flows. Critical flow is unstable and often sets up

standing waves between super and subcritical flow. When the actual depth is below the

critical depth, it is called supercritical because it is in a higher energy state. Likewise, if

actual depth is above critical depth it is called subcritical because it is in a lower energy

state.

2.5 Similitude Analysis

In the similitude analysis, the geometric and flow similarity requires the model

and the flow to be identical as the real model and flow. This is to ensure the result

19

obtained from the model study is well presented in order to predict full scale behavior. If

similarity has been obtained between model and prototype, the Froude number of the

model and the prototype must be the same for flow conditions where inertial and

gravitational forces are dominant.

For similarity of the flow patterns, the Froude number shall be equal in model

and prototype (American National Standards for Pump Intake Design, ANSI/Hydraulic

Institute Standards 9.8, 1998):

(2.3)

where,

Fr = Froude ratio parameter

Fm = Froude model parameter

Fp = Froude prototype parameter

A reasonable large geometric scale is selected to minimize viscous and surface

tension scale effects, and to reproduce the flow pattern in the vicinity of the intake. The

model also shall be large enough to allow visual observations of flow patterns, accurate

measurements of swirl and velocity distribution and sufficient dimensional control.

Froude number and Reynolds number for the model and prototype cannot be made

equal. Fixing the same Froude number for model and prototype results in the velocity

being reduced in the model depth and a fixed gravity constant. Fixing the Reynolds

number results in the velocity being increased in the model, given geometric scale

reduction of dimensions and constant kinematics viscosity.

Froude and Reynolds number equality could only be achieved simultaneously

by using a fluid with suitable kinematics viscosity in the model to adjust the Reynolds

number to match the prototype Reynolds number. However this is not possible and in

20

practice, water is used in the model as well as in the prototype. In the free surface

work, the gravitational forces are the most important value, so that the Froude number

must be made equal in the prototype in preference to the Reynolds number. This

ensures surface profiles, rotational flow and waves are correctly represented. For the

turbulent flow, the Reynolds number is not particularly important as long as both model

and prototype have values in the same flow regime (Abustan et al. 2004). If the

reduced Reynolds number of a model approaches the transitional point of turbulent to

laminar flow then laminar flow could occur in the model but turbulent or transitional flow

would occur in the prototype. Clearly this is not acceptable and consequently a

minimum operable Reynolds number has to be chosen (Abustan et al. 2004).

2.6 Mechanisms Involved in Vortex Formation

A number focused experimental and numerical investigations have provided

insight into the fundamental processes leading to the development of vortices in the

sump intakes. Shin et al. (1986) demonstrated that two basic mechanisms lead to inlet

vortex formation. The first mechanism involves the development of an inlet vortex due

to the amplification of ambient vorticity in the approach flow as vortex lines are

convicted into the inlet. The second mechanism involves the development of a trailing

vortex in the vicinity of the intake as a result of the variation in circulation along the

inlet. For this second case, a vortex can develop in a flow that is irrotational upstream,

and the vortex development therefore does not depend on the presence of ambient

vorticity. Shin et al. (1986) investigation on kinematic parameters, indicate that the

strength of an inlet-vortex or trailing vortex system increases with decreasing distance

from the surface. However, for an inlet in an upstream irrotational flow, two counter

rotating vortices can still trail from the rear of the inlet. Causes of vortex motion,

however, are still difficult to define for most practical situations.

21

2.7 Free Surface and Water Interface

Vortices in the vicinity of pump intakes may be adjacent to the channel bottom

or a channel wall (submerged vortices) or they may appear adjacent to the free surface

(free surface vortex). By studying the way that the vortex interact with the free surface,

more information can be known about the characteristics in vortex formation and

possibly better prediction methods can be developed as a result of such studies.

The study of the interactions with the free surface requires consideration of the

dynamics of the vorticity field bounded by a deformable surface. The surface deforms

to satisfy the conditions that the tangential stress is equal to zero and the normal stress

is equal to a constant at all times. According to Rood and Edwin (1995), the interaction

between vorticity and the surface characteristized by both the vorticity and flux of the

vorticity at the free surface, the deformation of the free surface, and the dynamic

behavior of the velocity field. The stresses at the interface between the two fluids

(water and air) must be in balance, such that:

(stress)water side of interface + (stress)air side + (stress)surface = 0 (2.4)

When describing the local details of flow at the free surface, viscous forces

normally cannot be neglected, especially when describing the vorticity generated by the

deformation. However, when localized flow details are not taken into account, there are

instances when the viscous terms can be neglected. By neglecting viscous forces, and

order of magnitude, estimates of the deformations can be obtained. For estimation of

the free surface deformation, the approximation is appropriate when the viscous force

is much less than the inertial force. This condition typically occurs when the flow is at

high Reynolds number.

22

2.8 Fundamentals of Vortex Flow in Sumps

Deviations in the pump approach flow distribution are the most common source

of swirl and vortex formation. Durgin and Hecker (1978) categorized the sources of

vortex formation into three types:

• Nonuniform approach flow to the sump due to the geometric orientation of sump

or approach channel or due to streaming flow patterns generated by

obstructions such as intake piers or columns.

• Existence of shear layers of high velocity gradients, including separated

boundary layers which are inherently rotational.

• Rotational wakes generated by objects or obstructions in the way of the

approach flow to the sump.

Padmanabhan (1987) reported that items first and second listed above are

major sources of vorticity in free surface and submerged vortices in pump sumps,

respectively. Similar, Anwar (1968) had mention that, the formation of vortices is

governed by two major factors which are submergence depth or distance from the

water surface to the entrance of the suction pipe and the swirl in the approach flow. He

also stated that a reduction in the swirl or an increase in submergence depth can

prevent the formation of vortices. Chang (1977) summarized the flow processes that

generate vortices in a pump sump as follow:

• Asymmetric approach flow – the approach flow has an inherent swirl which can

be magnified as the flow converges into the intake, due to the conservation of

angular momentum.

• Boundary discontinuities – changes in channel cross-section, diffusers, false

baffles, etc., can cause small eddies to shed, thus adding to the total vorticity.

The importance of this effect depends on how close the discontinuity is to the

23

intake and the strength of the eddy it creates. The presence of the intake itself

can also be a source of shading eddies.

• Boundary layer development – since the velocity at any solid boundary must be

zero because of viscosity, velocity gradients will be present in the boundary

layer which generate vorticity.

• Stagnation point flow – The flow at the plane surface near a gas turbine intake

showed the existence of a stagnation point towards which the boundary layer,

containing vorticity, flowed and was subsequently, drawn into the intake. Using

a combined boundary layer and potential flow analyses, a concentrating

stagnation point in the free surface is associated with vortex formation at

hydraulic intakes; and,

• Secondary layer - The presence of secondary flow currents in a plane

perpendicular to the main flow direction in straight rectangular channel. These

secondary flow velocities are only about 1% of the main velocity, but may be of

sufficient strength to contribute to the instability of vortices.

2.9 Problems Encountered in Pump Intakes

The various hydraulic problems associated with pump intakes include formation

of surface and subsurface vortices, prerotation and swirl, and flow separation at or near

the suction bell of either wet pit or dry-pit centrifugal pump. Any of these problems

adversely affect pump performance by causing cavitation, vibrations, and/or loss of

efficiency (Tullis, 1979; Alboleda and El-Fadel, 1996). Usually there is more than a

single reason for these problems, and the extents of the combined effects are seldom

predictable by mathematical modeling.

Formation of vortices, dependent on suction pipe velocity and submergence, is

strongly influenced by added circulation from vorticity sources, such as a nonuniform

24

approach flow resulting from intake and approach channel geometries; rotational

wakes shed from obstructions, such a columns or piers; and the velocity gradients

resulting from boundary layers at the walls and floor (Durgin and Hecker, 1978). The

circulation contributed by these vorticity sources is unpredictable and strongly depends

on intake design and operating conditions, especially for large pumping units with

multiple bays fed by a common approach channel. In these cases, physical modeling is

the only way of predicting the behavior of the prototype with a reasonable degree of

reliability.

2.9.1 Free Surface Vortex

Jain et al. (1978) reported that air entraining vortices, known as free surface

vortices, draw air into the intake from the water surface and thereby cause

considerable loss of pump efficiency and produce vibrations and noise. Free surface

vortices are most damaging to the pump when they draw air or trash into the intake

columns. The intensity of surface vortex varies directly with some functions of intake

and/or approach velocity and inversely with submergence, assuming upstream

conditions and other effects to be constant. The change in intensity is gradual, so the

specific point at which a vortex does or does not exist becomes a matter of definition

(Dicmas, 1987). To determine minimum submergence of the outlet pipe in the tank, the

Hydraulic Institute (1998) recommended the following relationship:

S = (1.0 + 2.3F) D (2.5)

where,

FD = Froude number

D = Diameter of inlet opening (m)

S = Submergence (m)

At certain stages of development, formations of vortices can be intermittent, so

practical definitions have to allow for these variations. Vortices in a model study usually

[ type thesis title here ]

Documents